AquaMILR: Mechanical intelligence simplifies control of undulatory robots in cluttered fluid environments

Abstract

While undulatory swimming of elongate limbless robots has been extensively studied in open hydrodynamic environments, less research has been focused on limbless locomotion in complex, cluttered aquatic environments. Motivated by the concept of mechanical intelligence, where controls for obstacle navigation can be offloaded to passive body mechanics in terrestrial limbless locomotion, we hypothesize that principles of mechanical intelligence can be extended to cluttered hydrodynamic regimes. To test this, we developed an untethered limbless robot capable of undulatory swimming on water surfaces, utilizing a bilateral cable-driven mechanism inspired by organismal muscle actuation morphology to achieve programmable anisotropic body compliance. We demonstrated through robophysical experiments that, similar to terrestrial locomotion, an appropriate level of body compliance can facilitate emergent swim through complex hydrodynamic environments under pure open-loop control. Moreover, we found that swimming performance depends on undulation frequency, with effective locomotion achieved only within a specific frequency range. This contrasts with highly damped terrestrial regimes, where inertial effects can often be neglected. Further, to enhance performance and address the challenges posed by nondeterministic obstacle distributions, we incorporated computational intelligence by developing a real-time body compliance tuning controller based on cable tension feedback. This controller improves the robot's robustness and overall speed in heterogeneous hydrodynamic environments.

Figures (7)

• Figure 1: The untethered mechanically intelligent limbless robot AquaMILR for undulatory locomotion in cluttered fluid environments. (A) The robot implements a decentralized bilateral actuation mechanism. Inset shows the design of the head module integrating power, computation, and communication modalities. (B) The robot navigates a laboratory model of an obstacle-rich environment.
• Figure 2: Programmable and quantifiable body compliance through bilateral cable actuation mechanism. (A) A geometric model illustrating a single joint, used to determine the exact lengths of the left and right cables ($\mathcal{L}^l$ and $\mathcal{L}^r$) necessary to achieve a specified joint angle ($\alpha$). (B) A schematic displaying various compliance states based on the generalized compliance variable $G$: bidirectionally non-compliant ($G = 0$), where the joint remains rigid and strictly follows the template trajectory (dashed line); directionally compliant ($G = 0.5$), where the joint allows movement that increases the angle (shown by the yellow region limited by upper and lower bounds indicated by blue and red lines); bidirectionally compliant ($G = 1$), where the joint permits movement in both directions with different degrees of flexibility; and fully passive ($G \geq 1.75$), where the joint can move freely in either direction. Figures adapted from wang2023mechanical.
• Figure 3: An example of robot locomotion in a regularly distributed lattice, showing frames of the robot's starting and ending poses, along with its tracked trajectory over time.
• Figure 4: The effect of the generalized compliance parameter ($G$) on locomotion performance. (A) The survivor function for varied $G$ values with respect to distance traveled, indicating the robot’s displacement before encountering an sticking pose or a motor failure. (B) Time-lapsed frames showing (i) an example of the robot becoming stuck at $G=0$ and (ii) an example of the robot successfully traversing the lattice at $G=1$.
• Figure 5: The effect of gait parameters on locomotion performance. (A) Success traversal rate as a function of spatial frequency ($\xi$). (B) Success traversal rate as a function of amplitude ($A$).